Science in the 20th century has given us a microscopic understanding of the world, ranging from the atomic realm of matter over the biological model of life to the psychology of individual thought and behavior. An important endeavor in the 21st century will be to understand the complex emergent patterns that characterize interacting systems in the physical, biological, engineering and social sciences.The vision of the AU Network for Computer Modelling of Complex Interactions is to combine expertise from physics, engineering, economics, business, cognitive neuroscience, and evolutionary biology to develop novel approaches for the study of intelligent, dynamically evolving systems.
All participants are established researchers in their fields who share a common methodological approach: combining simulation, experimentation, and statistical modeling in the study of small and large scale interactions in complex systems. One example where we expect our interdisciplinary cooperation to lead to crucial advances is in agentbased modeling: whereas physicists typically try to optimize quantum-physical processes with unrealistically simple rules for the interaction between agents, social scientists face the opposite challenge of limiting the rationality of their agents. The network participants share the vision to bridge this gap by modeling realistic agents, in between the extremes of the very primitive and the completely rational.
What is agent based moddeling?
Agent based modelling (ABM) is a computational framework for the study of interacting entities called agents, and the complex phenomena that emerges though their interactions. The framework is very flexible and means that ABM weaves it’s way though practically all scientific disciplines, from physics and engineering to economics and social sciences. The generality of agent based models is reflected, for instance, in the diverse backgrounds of the CMCI members.
There is no limit on what the agent is; a neuron, an ant, a human logged on to Facebook or an institution are all equally viable possibilities. Depending on their properties, the agents can react to their environment and their neighbours with a a simple mechanical response, or make complicated, deliberated decisions on their actions. As such, agent based models can be used to describe traffic flows, cognitive processes and viral videos in social media, just to give a few examples.
ABM provides an alternative viewpoint to the typical reductionist approach of science. Rather than explaining phenomena by breaking it down to the constituents, ABM shows that often things are more than the sums of their parts. Although increasing rapidly in popularity, ABM is still in the phase of rapid growth, both in terms of developing good modelling schemes, and in terms of applying them to concrete models.